A Deep Classification System for Medical Data Analysis

作者：Hassan Ayah M^*; Hassan Yasser F; Kholief Mohamed H

来源：Journal of Medical Imaging and Health Informatics, 2018, 8(2): 250-256.

DOI：10.1166/jmihi.2018.2293

• PSO-ANN
• Breast Cancer Classification
• Deep Learning

Medical records encompass entities representing patients' vital aspects. Whenever incorporated, entities assist in building up a comprehensive diagnosis of the patient. However due to the complicated and interlinked nature of the large-scale medical data, certain systematic procedures have to be followed for the proper and efficient classification. These are not only confined to the classification process but they also extend to the preprocessing steps. Preprocessing steps include segmentation of unwanted regions in images, selection of features for big data and degrading textual reports to their abstract form. Regardless of the fact that this model can be used for different types of medical data, it is highly suitable for cancer medical data. This is due to the fact that cancer medical data contain different datasets; namely; numbers, images and text medical reports. The Particle Swarm Optimization (PSO) and Artificial Neural Networks (ANN) were used to develop an algorithm to classify breast cancer datasets. The algorithm effectively differentiated between malignant and cancer free datasets with a 93.2% accuracy. The model provides an efficient, tool for processing breast cancer datasets.

• 出版日期2018-2